Mechlorethamine: The First Approval of a Cancer Drug and What "Approval" Guaranteed in 1949 (1949)
In this article
Mechlorethamine: The First Approval of a Cancer Drug and What "Approval" Guaranteed in 1949 (1949)
On day 120 we discussed the 1946 publication on nitrogen mustard, and on day 123 where a laboratory's money came from. Today we look at one date: 15 March 1949. In news articles, mechlorethamine (trade name Mustargen) is cited as the first chemical drug to receive official approval from the Food and Drug Administration (FDA) for cancer treatment on that date. But what did "approval" guarantee that year? The answer may be more modest than the one that comes to mind today.
Reading a drug from its label
On my desk is a label for mechlorethamine. It is dated 2004; hemonc.org reports that the oldest version found in Drugs@FDA is dated 9 February 2004. So this is a text from half a century after 1949, not from 1949 itself. Even so, few documents state the nature of a drug so plainly.
The label describes the drug as "highly toxic" and asks that it be given under the supervision of a physician experienced in cancer chemotherapy. The drug is "a powerful vesicant": if it leaks outside the vein, painful inflammation, hardening and tissue loss of the skin may follow. Nausea and vomiting are dose-limiting; vomiting begins one to three hours after the dose. The lymphocyte count falls within 24 hours, the granulocyte count drops markedly around days 6-8, and this fall can last from ten days to three weeks. The total dose of a course is limited to 0.4 mg per kilogram of body weight. Staff who prepare it are asked to work in a biological safety cabinet, with gloves, goggles and gown. The indications on the label all begin with the word "palliative": stage III-IV Hodgkin's disease, lymphosarcoma, chronic leukemias, polycythemia vera, mycosis fungoides, bronchogenic carcinoma, and intracavitary use in widespread cancers with fluid accumulation.
Someone reading this label does not draw an expectation of a cure from the phrase "the first approved cancer drug". Nor should they. To understand the 1949 approval, one has to look at the legal ground of that year.
The first approval on record
A news article published by MDedge, "First FDA-approved chemo agent turns 60", states that on 15 March 1949 mechlorethamine became the first chemical agent to receive FDA approval for treating cancer. hemonc.org gives the same date and calls the drug the first approved chemotherapy drug. Both sources are secondary; I did not see the FDA's document of that day or the application file. Which cancer it was approved for, who the applicant was and what the file contained are not in the sources I read. The name Merck on the label belongs to the 2004 version; I do not know who the applicant was in 1949, so I do not write it.
The background was told in earlier days of the series: the animal experiments of Goodman and Gilman at Yale and the first patient (day 115), the results kept secret in the 1940s, and the 1946 publication (day 120). Briefly, then: according to DeVita and Chu the first study was carried out in 1943, the results were not published until 1946 because of the secrecy around mustard gas, and the responses turned out to be "brief and incomplete". The JAMA paper by Goodman, Wintrobe, Dameshek, Goodman, Gilman and McLennan (1946; 132(3): 126-132) covered Hodgkin's disease, lymphosarcoma, leukemia and related disorders.
What did approval measure?
The FDA's own history page summarizes the Federal Food, Drug, and Cosmetic Act, signed on 25 June 1938, like this: for the first time, manufacturers were required to show that a drug was safe before marketing it. An application was filed; if the FDA did not act within a set period, the application became approved automatically. The law had no requirement to prove effectiveness. That requirement came in October 1962 with the Kefauver-Harris Amendments: firms now had to present "substantial evidence" based on "adequate and well-controlled studies" of effectiveness for the intended use, and the FDA had to approve the application explicitly. The same amendments also introduced informed consent from people taking part in studies; that is, the Nuremberg principle we discussed on day 121 entered American drug law fifteen years late.
From this I draw an inference, and I say it is an inference: the legal standard of 1949 was safety; so reading 15 March 1949 as the date of what we today mean by "approval", proof of efficacy, would be wrong. I infer this from the text of the law; I did not see the inside of the 1949 file, so I do not know what the FDA examined that day. And "safe" here does not mean "harmless": the same drug's label half a century later describes it as "highly toxic".
Three separate questions: is it safe, does it work, for whom and how much?
"Approval" is not one answer to one question. The first is the safety question: is there an unacceptable harm in the stated use? The second is the efficacy question: does the drug really work, and on what evidence? The third is the placement question: for which patients, in what order, alongside which alternatives? The 1938 law rests on the first question, the 1962 amendments on the second; the third is shaped long after the approval decision, by clinical guidelines and years of experience.
The same molecule's situation in 2013 shows the difference. Mechlorethamine gel (Valchlor) was approved in August 2013 for stage IA and IB mycosis fungoides-type cutaneous T-cell lymphoma in patients who had already received skin-directed therapy. The basis of the application was a randomized, observer-blinded non-inferiority trial of 260 patients in 13 centers. The new gel was compared with a pharmacy-compounded mechlorethamine ointment; at month 12 the confirmed response was 60 percent with the gel and 48 percent in the comparison arm (a 12-point absolute and 1.25-fold relative difference). This figure does not mean "mechlorethamine works in 60 percent of patients": two mechlorethamine preparations were compared, not a placebo. Even so, putting what a drug was approved on in 2013 and in 1949 side by side shows the distance between them.
Cost and limits: a brief response, a drug hard to handle
First, the duration of the response. DeVita is quoted as saying that nitrogen mustard "worked, but only very briefly" (interview at the Cancer History Project; I give it as a conceptual summary). According to DeVita and Chu these brief and incomplete responses led to the pessimism of the 1950s; even those who developed the drug became, after a while, among the sharpest critics of chemotherapy (William Dameshek was one of them). The MDedge article, attributing it to Franco Muggia of NYU, reports that in those years such drugs were "highly experimental" and used as a "last resort".
Second, the drug's own cost. The label states that in animal studies the drug is corrosive, carcinogenic, mutagenic and teratogenic; bone-marrow suppression can be severe and, in the label's wording, an overdose can be life-threatening. This information is from the 2004 label; I do not know to what extent it was known in 1949. The protective measures for the staff who prepare it are a reminder that the drug is a derivative of sulfur mustard, once used as a chemical weapon (The Hospitalist); that mustard gas had also appeared, in the Bari harbor incident we described on day 116, with suppression of lymphocytes and bone marrow.
Third, a single-sourced "first". The phrase "the first approved cancer drug" rests on two secondary sources; I did not verify it in the FDA's own record list. That is why I built the sentence as "cited in the sources as the first".
What remains
- The lineage of alkylating drugs: interest extended to chlorambucil and then cyclophosphamide (DeVita and Chu)
- The continuing place of mechlorethamine in diseases such as Hodgkin's disease and mycosis fungoides
- The idea that a drug being "approved" and being "proven" are separate questions
What fell or was limited
- The assumption that approval meant proof of efficacy (not a legal requirement until 1962)
- The idea of single-drug, palliative, short-lived treatment
- Reading the sentence "first approval" as an unqualified fact (secondary source, not verified)
One molecule, two standards. The axis covers 1936-2014; the tick marks are ten-year intervals. Orange boxes show thresholds in drug law (1938: safety requirement; 1962: efficacy requirement), green boxes show the history of mechlorethamine (1949 approval; 2013 gel approval). The intermediate dates (1943 first patient, 1946 publication) are not on the diagram because they were told on earlier days.
License, evidence and the patient
I could not find a verified source on how mechlorethamine or this approval date was reflected in Türkiye; I am not inventing a local tie. I can write a general inference: a drug being licensed does not by itself say on what evidence, in which group of patients and against which alternatives it was assessed; licensing standards change from country to country and year to year. So for a patient or relative, the useful question is as much "against what, in how many patients, with what result?" as "is it approved?".
From 1938 to today
What was confirmed, what conflicts
- Date of approval: 15 March 1949 (MDedge; hemonc.org, "1949-03-15"). Both are secondary; one may derive from the other. I did not see the FDA's own 1949 record or a Drugs@FDA entry.
- "First": MDedge "the first chemical agent to receive FDA approval for cancer treatment"; hemonc.org "the first approved chemotherapy drug". Not verified in the FDA's historical approval list; hedged in the text as "cited in the sources".
- 1949 indication and applicant: not in the sources I read. MDedge does not say which cancer it was approved for. The list of indications on the label and the name Merck belong to the 2004 version and were not used for 1949.
- Date of the first patient: MDedge "early December 1942, Yale"; DeVita and Chu "1943". The details and the dispute over the date are on day 115; not important here, and not put on the diagram.
- JAMA paper: Goodman LS, Wintrobe MM, Dameshek W, Goodman MJ, Gilman A, McLennan MT. JAMA 1946;132(3):126-132 (James Lind Library). The date 21 September 1946 is from The Hospitalist's reference list. I did not read the full text or the patient numbers; I did not use them.
- 1938 law: signed 25 June 1938; manufacturers were for the first time required to show safety; if the FDA did not act within a set period the application became approved automatically (FDA history page). The number of days is not on that page; I did not write it. The conclusion that there was no efficacy requirement is an inference from how the law is described.
- 1962 amendments: October 1962; "substantial evidence of effectiveness", "adequate and well-controlled studies", explicit FDA approval, informed consent of study subjects (FDA history page).
- Label: hemonc.org gives the oldest label in Drugs@FDA as 9 February 2004. The PDF I read (Merck, FDA MedWatch archive) is a 2004 label; its copyright line is 1985 and 1999, and its "issued" line is blank. The toxicity and dose information is from this text; it does not count as 1949 information.
- Muggia's remark: attributed in the MDedge article to Franco Muggia of NYU; given as a conceptual summary. DeVita's "it worked very briefly" is from the Cancer History Project interview; again a conceptual summary.
- Valchlor: The news item is dated 26 August 2013 (Practical Dermatology); the exact approval day is not in it, so "August 2013" is written. Pivotal trial: 260 patients, 13 centers, randomized, observer-blinded, non-inferiority; 60 percent versus 48 percent (month 12, confirmed response). The trial enrolled stage IA-IIA patients; the approved indication is stage IA-IB. The comparison arm was not a placebo but a pharmacy-compounded mechlorethamine. The 12-point and 1.25-fold differences drawn from 60 and 48 percent are my own calculation.
- Out of scope: the content of the 1949 file, how many patients were reviewed, what data the FDA looked at. These could not be answered in this piece.
Doll, Hill, Wynder and Smoking (1950)
Who first showed the link between smoking and lung cancer, by what method and on what date? In the same year more than one study asked this question, and each measured something different. Tomorrow I will tell which of them can be counted as "first" and by what yardstick, once dates and sources are verified.
Author's note: The sentence "the first approved cancer drug" sounds like a milestone. I wanted to turn it into a question: what did approval guarantee that year? Part of the answer is an inference from the text of the law, and part rests on secondary sources; I did not see the 1949 file itself. I have separated this in the fact-check box.
Sources and Further Reading
- MDedge / The Hospitalist. First FDA-approved chemo agent turns 60 (mdedge.com/internalmedicine/article/15283/oncology/first-fda-approved-chemo-agent-turns-60).
- HemOnc.org. Mechlorethamine (Mustargen) (hemonc.org/wiki/Mechlorethamine_(Mustargen)).
- U.S. Food and Drug Administration. Promoting Safe and Effective Drugs for 100 Years (fda.gov/about-fda/histories-product-regulation/promoting-safe-effective-drugs-100-years).
- Merck & Co. MUSTARGEN (mechlorethamine HCl for injection), prescribing information, 2004; FDA MedWatch archive copy.
- Goodman LS, Wintrobe MM, Dameshek W, Goodman MJ, Gilman A, McLennan MT. Nitrogen mustard therapy. JAMA 1946;132(3):126-132 (citation from the James Lind Library; full text not read).
- DeVita VT Jr, Chu E. A history of cancer chemotherapy. Cancer Res 2008;68(21):8643-8653 (only the opening section was read).
- Cancer History Project. Vince DeVita on the History of Chemotherapy (cancerhistoryproject.com).
- Practical Dermatology. FDA Approves Valchlor Gel for the Treatment of Stage IA and IB Mycosis Fungoides-Type CTCL, 26 August 2013.
- Mukherjee S. The Emperor of All Maladies: A Biography of Cancer. New York: Scribner, 2010. (Not separately verified for this piece.)
This piece is day 124 of the "365 Days of Cancer History" series and recounts the 1949 approval of mechlorethamine as a historical event. It is popular history; it is not a diagnostic or treatment recommendation. Consult your physician for decisions about your health. Sources have been verified; details that rest only on a secondary source are marked in the fact-check notes.
Was this article helpful?
Your feedback helps us improve our content.
